Self-triggered Model Predictive Control for Trajectory Tracking of Mobile Robot

Abstract

This paper considers self-triggered model predictive control (MPC) for trajectory tracking of a mobile robot with uncertain slips between the wheels and ground over a network. Since the trajectory tracking model of the robot has a time-varying term due to the uncertain slip, the paper introduces a reachable set of the trajectory tracking nominal model to construct tube constraints. The proposed tube-based MPC problem has the properties that the tracking error and control input satisfy constant constraints under the uncertainty and that the MPC problem has recursive feasibility. Furthermore, the paper employs a self-triggered mechanism to cope with a remote manipulation of the robot over the network and reduce the communication. Numerical examples illustrate the effectiveness of the proposed tube-based MPC.